Dendrites of the dorsal and ventral hippocampal CA1 pyramidal neurons of singly housed female rats exhibit lamina-specific growths and retractions ...

Page created by Marshall Holmes
 
CONTINUE READING
Received: 26 February 2018    |   Revised: 27 March 2018   |   Accepted: 1 April 2018

DOI: 10.1002/syn.22034

RESEARCH ARTICLE

Dendrites of the dorsal and ventral hippocampal CA1 pyramidal
neurons of singly housed female rats exhibit lamina-specific
growths and retractions during adolescence that are responsive
to pair housing

Yi-Wen Chen1 | Ada Akad1 | Ruka Aderogba1 | Tara G. Chowdhury1 |
Chiye Aoki1,2

1
 Center for Neural Science, New York
University, New York, New York 10003
                                                   Abstract
2
 Neuroscience Institute, Langone Medical           Adolescence is accompanied by increased vulnerability to psychiatric illnesses, including anxiety,
Center, New York University, New York,             depression, schizophrenia, and eating disorders. The hippocampus is important for regulating emo-
New York 10016                                     tional state through its ventral compartment and spatial cognition through its dorsal compartment.

Correspondence                                     Previous animal studies have examined hippocampal development at stages before, after or at sin-
Chiye Aoki, PhD, 4 Washington Place,               gle time points during adolescence. However, only one study has investigated morphological
Room 809, New York, NY 10003.                      changes at multiple time points during adolescence, and no study has yet compared developmental
Email: ca3@nyu.edu
                                                   changes of dorsal versus ventral hippocampi. We analyzed the dorsal and ventral hippocampi of
Present address
                                                   rats to determine the developmental trajectory of Golgi-stained hippocampal CA1 neurons by sam-
Tara G. Chowdhury, Department of Behav-
ioral Neuroscience, Oregon Health and Sci-         pling at five time points, ranging from postnatal day (P) 35 (puberty) to 55 (end of adolescence).
ence University School of Medicine, Portland,      We show that the dorsal hippocampus undergoes transient dendritic retractions in stratum radia-
Oregon
                                                   tum (SR), while the ventral hippocampus undergoes transient dendritic growths in SR. During

Funding information                                adulthood, stress and hormonal fluctuations have been shown to alter the physiology and mor-
The Klarman Family Foundation Grant                phology of hippocampal neurons, but studies of the impact of these factors upon adolescent
Program in Eating Disorders Research,              hippocampi are scarce. In addition, we show that female–female pair housing from P 36–44 signifi-
National Institutes of Health, Grant
                                                   cantly increases branching in the dorsal SR and reduces branching in the ventral SR. Taken
Number: R21MH091445-01; Grant Num-
ber: R21 MH105846; Grant Number:                   together with data on spine density, these results indicate that pyramidal cells in the dorsal and
R01NS066019-01A1; Grant Number:                    ventral CA1 of female adolescents are remodeled differently following single housing. Social hous-
R01NS047557-07A1; Grant Number: NEI
                                                   ing during adolescence elicits pathway-specific changes in the hippocampus that may underlie
Core Grant EY13079; NYUs Research Chal-
lenge Fund, Grant Number: NSF-REU                  behavioral benefits, including stability of emotion regulation and superior cognition.
1460880; the Fulbright Scholarship;
National Center for the Advancing of
Translational Science (NCATS), Grant               KEYWORDS
Number: UL1 TR000038; National Insti-              apical dendrites, dendritic spines, Golgi stain, pair housing, Sholl analysis
tutes of Health, Grant Number:
R25GM097634-01

1 | INTRODUCTION

Adolescence is a critical period of life, characterized by maturation of cognitive, reproductive and social skills and capacities in all mammals (Hazen,
Schlozman, & Beresin, 2008; Sisk & Foster, 2004). Adolescence is also the most common stage for the emergence of neuropsychiatric illnesses, such
as schizophrenia, depression, substance abuse, anxiety disorders, and eating disorders (Kessler et al., 2007; Paus, Keshavan, & Giedd, 2008). Previous
studies suggest that abnormal trajectories in brain development during adolescence may contribute to the pathophysiology of these disorders (Meri-
kangas, Nakamura, & Kessler, 2009; Paus et al., 2008).
    There is mounting evidence indicating that increases in circulating steroids during adolescence, which accompany reproductive maturation, influ-
ence the development of neurons in the hippocampus and other brain regions (Cooke & Woolley, 2005; Giedd et al., 2006). Imaging studies of

Synapse. 2018;e22034.                                            wileyonlinelibrary.com/journal/syn                    V
                                                                                                                       C 2018 Wiley Periodicals, Inc.   |   1 of 17
https://doi.org/10.1002/syn.22034
2 of 17   |                                                                                                                                  CHEN   ET AL.

healthy brains indicate that the human brain undergoes dynamic changes throughout adolescence and that the change in gray matter volume over
time has an inverted U-shape pattern (Casey, Jones, & Hare, 2008; Giedd, 2004; Gogtay et al., 2004). The loss of gray matter in the prefrontal cortex,
temporal lobe, and basal ganglia is surmised to be a result of massive pruning of synapses and development of the white matter, reflecting the growth
of myelin that is delayed, relative to axonal growth (Casey, Tottenham, Liston, & Durston, 2005; Gogtay et al., 2004). Rodent models also shed light
upon our understanding of healthy brain development during adolescence. Some specific examples include overproduction and rapid pruning of den-
dritic branches, dendritic spines, synapses, and/or neurotransmitter expression during adolescence in the amygdala (Zehr, Todd, Schulz, McCarthy, &
Sisk, 2006), nucleus accumbens (Teicher, Andersen, & Hostetter, 1995) and prefrontal cortex (Andersen & Teicher, 2004; Andersen, Thompson, Rut-
stein, Hostetter, & Teicher, 2000; Drzewiecki, Willing, & Juraska, 2016; Koss, Belden, Hristov, & Juraska, 2014; Willing & Juraska, 2015).
    As for the hippocampus, adolescents perform differently from adults on hippocampal-dependent learning and memory tasks (McCormick &
Mathews, 2010). The rate of cell proliferation in the dentate gyrus of the hippocampal formation, which is the highest at 2 weeks postnatal for
rodents, declines progressively but is still higher during adolescence than in adulthood (Bayer & Altman, 1974; He & Crews, 2007; Schlessinger,
Cowan, & Gottlieb, 1975). Spines are pruned during adolescence (Afroz, Parato, Shen, & Smith, 2016) and this pruning ensures optimal
hippocampus-dependent spatial memory (Afroz et al., 2016). However, it remains unexplored whether remodeling of hippocampal circuit during ado-
lescence involves dendritic branch remodeling as well as spine pruning.
    The cortical and limbic regions that continue to mature during adolescence, such as prefrontal cortex, amygdala, and hippocampus, are also
some of the most stress-reactive areas in the brain (McEwen, 2005). The hippocampus, in particular, shows plasticity in its response to stress (Con-
rad, Ortiz, & Judd, 2017; McEwen, 1999) and gonadal hormones (Woolley & McEwen, 1992) independently and jointly (McLaughlin et al., 2010).
The hippocampus is not a functionally uniform structure. Instead, function differs along the septo-temporal axis of the hippocampus, with the dorsal
hippocampus being more important for spatial learning and memory performance and with the ventral hippocampus being more preferentially
involved in regulation of stress, emotion, and affect (Bannerman et al., 2002; Fanselow & Dong, 2010; Moser, Moser, & Andersen, 1993). Accumu-
lated evidence suggests that the dorsal versus ventral hippocampus also differ in cell morphology, topographical arrangement of neuronal connectiv-
ity, and patterns of gene expression (Fanselow & Dong, 2010; Swanson & Cowan, 1977). Previous findings from our laboratory add to this list of
differences: there is a dorsal-ventral distinction in the changes in dendritic branching that are elicited by food-restriction stress during adolescence
(Chowdhury et al., 2014a). Adolescence is a period of psychiatric vulnerability which, in turn, is influenced by anxiety that is proposed to be regu-
lated more by the ventral than the dorsal hippocampus (Bannerman et al., 2002; McHugh, Deacon, Rawlins, & Bannerman, 2004). However, we
have data indicating otherwise (Aoki, Chen, Chowdhury, & Piper, 2017). It has also been shown that dorsal and ventral hippocampal lesions interact
with puberty to result in different behavioral deficits in adulthood (Lipska, Jaskiw, & Weinberger, 1993; Pacteau, Einon, & Sinden, 1989). However,
no study has examined or directly compared changes in the structure of pyramidal neurons in the dorsal and ventral hippocampus during the devel-
opmental period specifically spanning adolescence.
    Environmental enrichment and social experience have been shown to alter both hippocampal-dependent behaviors and neuronal structure in
the hippocampal CA1 (Faherty, Kerley, & Smeyne, 2003; Huttenrauch, Salinas, & Wirths, 2016; Lauterborn, Jafari, Babayan, & Gall, 2015). In both
animals and humans, social interaction is important for parental care, pair bonding and cooperation (Clark & Dumas, 2015; Rilling & Young, 2014;
Trezza, Campolongo, & Vanderschuren, 2011). Absence of exposure to age-appropriate social interactions can be detrimental to the organism and
the gene pool. For humans, social isolation during adolescence (both voluntary and involuntary) is associated with several neuropsychiatric illnesses,
including schizophrenia, anxiety, and depression (Rubin, Coplan, & Bowker, 2009). Adolescent rodents also live in groups and exhibit higher levels of
social behaviors than younger or adult animals (Panksepp et al., 2007). Therefore, rodent models of adolescent social stress can be used for under-
standing the impact of social stress in adolescent humans.
    Most studies involving laboratory rodents have focused on changes elicited within brains of males, leaving female brains to be relatively under-
studied. Few studies have found that exposure to enriched rearing environment from weaning leads to sex differences in dendritic morphological
changes in dentate gyrus (Juraska, Fitch, Henderson, & Rivers, 1985) and hippocampal CA3 region (Juraska, Fitch, & Washburne, 1989). Previous
studies also suggest that females have an increased sensitivity to chronic social stress, and that social instability stress reduces hippocampal neuro-
genesis in female adolescent rats (McCormick, Nixon, Thomas, Lowie, & Dyck, 2010). Sex differences in behaviors (Galef & Sorge, 2000) and corti-
costerone (Hurst, Barnard, Hare, Wheeldon, & West, 1996) responses to housing conditions have been reported for rats as well. Although
adolescents exhibit heightened levels of social behavior and social isolation alters neurotrophin levels in the hippocampus (Meng, Li, Han, Shao, &
Wang, 2011; Scaccianoce et al., 2006), it remains unexplored whether deprivation of social interaction due to single housing or the experience of
pair housing during adolescence impact dendritic morphology of neurons in the female hippocampus.
    In this context, our present set of experiments had three objectives. First, we measured dendritic complexity and spinous complexity in female
rats throughout adolescence, to determine the developmental trajectory of the hippocampal CA1 neurons. Secondly, we investigated whether the
dorsal and the ventral hippocampi show similar or different patterns of development. Finally, we evaluated the contribution of social housing during
early to mid-adolescence. The analysis reveals that dendritic arbor and spines of pyramidal cells in the hippocampal CA1 are remodeled during mid-
adolescence in ways that differ for the dorsal versus ventral hippocampus and are influenced by social housing. Together, our findings contribute to
our basic understanding of how wiring of the hippocampus changes across adolescence and shed light on the mechanisms of the behavioral benefits
derived from social housing during adolescence.
CHEN   ET AL.                                                                                                                               |   3 of 17

2 | EXPERIMENTAL PROCEDURES

2.1 | Animals
All procedures relating to the use of animals were in accordance with the Institutional Animal Care and Use Committee of New York University (Ani-
mal Welfare Assurance #11–1374).
    Two cohorts, totaling 36 female Sprague-Dawley rats, were purchased from Hilltop Lab Animals INC and delivered to the New York University
animal facility on postnatal Day 28 (P28). All animals were handled and weighed daily, kept on a 12-h light:12-h dark cycle (lights off at 19:00 h),
and given food and water ad libitum until they were sacrificed as described below under ‘Brain collection and tissue processing’.
    Twenty of these animals (cohort 1) were used to analyze dendritic development of hippocampal CA1 neurons. Upon arrival, these animals were
pair-housed with another female. Starting on P32, animals were singly housed and assigned to one of five age groups, to be euthanized either at
P35, P40, P44, P50, or P55. Each animal was weighed on P32, prior to grouping, so as to equalize the mean body weights across the age groups.
Each animal was weighed daily.
    The second cohort of 16 rats was used to determine the effects of pair housing during adolescence upon dendritic remodeling. Upon arrival,
these animals were weighed, then pair-housed with another female. On P36, animals were assigned into two groups of eight animals each: singly-
housed or pair-housed with the same female. The only criterion used to house the animals singly or as a pair was their body weights: the mean
weight of the two groups was equalized. All animals of the second cohort were weighed daily and euthanized on P44.
    While it is known that ovarian hormones affect hippocampal structure and function, we did not consider the estrous cycle phase as a biological
variable, because it has been shown that pubescent female Sprague-Dawley rats (P35–41) exhibit only partial cycling (Hodes & Shors 2005). Vaginal
smears of animals at P50 indicated that two were in diestrus, and the other two were in transition from estrus to metestrus. At P55, two were in
diestrus, one was in transition from proestrus to estrus and the fourth was in metestrus.

2.2 | Brain collection and tissue processing
Animals in the first cohort were euthanized by being deeply anesthetizing with urethane (34%; 0.65–0.85 mL/g body weight, i.p.) between the hours
of 12–4 pm, during the light cycle. The animals were decapitated, and the brains were quickly removed from the skull. The brain was divided along
the coronal plane into three blocks of 2–3 mm thickness and processed immediately for Golgi–Cox impregnation using the FD Rapid GolgiStain kit,
according to the instructions of the manufacturer (FD NeuroTechnologies, Ellicott City, MD).
    Animals in the second cohort were euthanized on P44 between the hours of 4–6 pm, at the end of the light cycle. Animals were deeply anes-
thetized using urethane (34%; 0.65–0.85 mL/g body weight, i.p.) prior to transcardial perfusion with phosphate-buffered saline containing heparin
(10,000 U per 500 mL, Henry Schein) followed by 500 mL of 4% paraformaldehyde in 0.1M phosphate buffer (pH 7.4) at a flow rate of 50 mL/min.
The right hemisphere was divided along the coronal plane into three blocks of 2–3 mm thickness and processed immediately for Golgi–Cox impreg-
nation using the FD Rapid Golgi Stain kit according to the instructions of the manufacturer. The left hemisphere was reserved for another study
involving immunocytochemistry.
    For all brains collected from cohort 1 and 2, coronal sections, 250 lm thick, were prepared, using the Leica VT1000M Vibratome (Leica Micro-
systems GmbH, Wetzlar, Germany).

2.3 | Morphological analysis
CA1 pyramidal neurons were chosen from two compartments along the septotemporal axis of the hippocampus, namely rostral–dorsal, and caudal–
ventral hippocampal regions, heretofore referred to as ‘dorsal’ and ‘ventral’, respectively. Dorsal cells were taken from coronal sections correspond-
ing to 22.80 to 24.30 mm in the anterior–posterior axis from Bregma, and 2.50 to 3.50 mm deep from the skull surface. Selected neurons from
ventral hippocampus were chosen from coronal sections corresponding to 24.80 to 26.04 mm in the anterior–posterior axis from Bregma, and
7.00–9.40 mm deep from the skull surface, using the rhinal fissure as a local landmark (Figure 1a,b) (Chowdhury et al., 2014a). Golgi-impregnated
neurons were filled completely, such that pyramidal cell body, apical, and basal dendrites, and dendritic spines were clearly visible. Experimenters
traced and analyzed the neurons while kept blind to the identity of the experimental groups of the animals. Cells to be traced were chosen such
that their dendritic processes were not artificially broken within 300 lm of radius from the soma and were minimally overlapping with neighboring
pyramidal cells. The cell body and apical dendrites of cells were traced in three dimensions using the Neurolucida system 11.07 (MicroBrightField
Bioscience, Williston, VT) attached to an Olympus BX51 microscope (Olympus, Tokyo, Japan), using a 203 objective. Neurolucida Explorer software
package was used to analyze the manual tracing of the neuron using the ‘Sholl Analysis’ option to quantify the number of intersections that the api-
cal dendrite and its branches made with imaginary spheres centered at the center of the soma, beginning at a radius of 20 lm and increasing in dis-
tance by 20 lm (Chowdhury et al., 2014a,b). In cohort 1, Sholl analysis was performed on 12–20 cells per group from four animals (4–5 cells per
animal). In Experiment 2, Sholl analysis was performed on 30–37 cells for both groups, with each group consisting of eight animals (3–5 cells per
animal).
4 of 17   |                                                                                                                               CHEN   ET AL.

FIGURE 1       Schematics showing the outline of typical coronal sections selected for sampling CA1 pyramidal cells of the rostral–dorsal
[‘dorsal’, (a)] and caudal–ventral [‘ventral’ (b)] hippocampus; see “Methods” for further details of the range of coordinates used to sample
from the two compartments of the hippocampus. The inserts show orientations of pyramidal neurons and two distinct laminas of CA1,
stratum radiatum (SR) and stratum lacunosum-moleculare (SLM). Calibration bar represents 2 mm for the hemispheres and 200 mm for the
neurons

    Spine density and percentage of mature spines were also analyzed. At least five segments of dendrites were sampled, consistently within stra-
tum radiatum (SR). Because pyramidal neurons are undergoing overall growth during adolescence, the distance of SR from soma varied with age.
The chosen dendritic branches occurred between 80 and 160 mm from the soma for brains collected at P35; 100 and 200 mm from the soma for
brains collected at P40 and P44; 120 and 220 mm from the soma for brains collected at P50 and P55. Within SR, the segments were chosen to be
near where the number of intersections made by the dendritic branches with imaginary spheres in the Sholl analysis were maximum for each age
group.
    We also analyzed spine density in stratum lacunosum-moleculare (SLM). Here, too, the distance of SLM from soma varied with age. Thus, the
chosen dendritic branches occurred 260 and 360 mm from the soma for brains collected at P35; between 300 and 400 mm from the soma for brains
collected at P40 and P44; between 340 and 440 mm from the soma for brains collected at P50 and P55). Five dendritic segments from SLM were
analyzed per brain. In SLM, the segments were chosen to reside within 100 mm from the most distal imaginary sphere of the Sholl analysis.
    For both SR and SLM, dendritic segments were traced using the Neurolucida software along with dendritic spines, using a 1003 objective. Den-
dritic spines were categorized based on their morphological characteristics into four groups: thin (immature), filopodia (immature), stubby, and mush-
room spines. The relative proportion of mature spines was calculated for each dendritic segment as the ratio of the density of mushroom and
stubby spines to the total spine density (Chowdhury et al., 2014a,b).

2.4 | Statistical analysis
All statistical analyses were conducted, while keeping the experimenter blinded to the environmental condition and the animal’s age at the time of
euthanasia. Data from the Sholl analysis were analyzed using a repeated measure two-way ANOVA to reveal the overall group effect. Two-tailed
Student’s t test was used to assess the difference when two groups were being compared. Two-way ANOVA was used to evaluate the significance
of differences among the five age groups and between the two hippocampal regions, using postnatal days and dorsal or ventral hippocampal CA1 as
the two factors. One-way ANOVA was used when three or more groups were being compared within each hippocampal CA1 region. Significant
pairs were determined using Tukey’s HSD post hoc test. For all analyses, neuron was regarded as the independent unit of analysis (N value), and we
pooled neurons from multiple animals of the same age and rearing condition. Detection of outliers was performed for each group, using the ROUT
method (Motulsky & Brown, 2006). This method led to the detection of two values, no animal contained cluster of outliers, and the group mean
CHEN   ET AL.                                                                                                                                |   5 of 17

values was not altered by excluding the outliers. Thus, the data are presented without removing the outliers. Statistical analysis was done by SPSS
statistical software 24.0 (IBM Corp., Armonk, NY) and GraphPad Prism 7.0. p values of
6 of 17   |                                                                                                                        CHEN   ET AL.

FIGURE 2     Apical dendrites of the dorsal and ventral hippocampal CA1 neurons exhibit growth and retraction during adolescence. (a) CA1
pyramidal cells of the dorsal hippocampus transiently decrease in complexity during adolescence. Representative Neurolucida tracings of
dorsal hippocampal CA1 pyramidal cells from each age point. Calibration bar represents 100 mm. The average number of dendritic
intersections and the total dendritic length in the SR and SLM were calculated for each cell within each age group. In dorsal hippocampal
CA1, comparisons of dendritic branching across age groups revealed that the average number of dendritic intersections in the SR (b), and in
the SLM (d) transiently decrease during adolescence. Comparisons of dendritic branching across age groups also revealed that the total
dendritic length in the SR (c), and in the SLM (e) transiently decrease during adolescence. (f) CA1 pyramidal cells of the ventral hippocampus
transiently increase in complexity during adolescence. Representative Neurolucida tracings of ventral hippocampal CA1 pyramidal cell from
each age point. Calibration bar represents 100 mm. In ventral hippocampal CA1, comparisons of dendritic branching across age groups
revealed that the average number of dendritic intersections in the SR (g), and in the SLM (i) transiently increase during adolescence.
Comparisons of dendritic branching across age groups also revealed that the total dendritic length in the SR (h), and in the SLM (j)
transiently increase during adolescence. Data from each group were compared across ages from P35 to P55. Bar graphs represent
means 1 SEM (N 5 15–20 cells per group). Same letters indicate significant differences between age groups as assessed by one-way
ANOVA, followed by Tukey’s HSD test. Same letters with the prime symbol indicate marginally significant differences (p < .08) between
groups as assessed by Student’s t tests
CHEN   ET AL.                                                                                                                               |   7 of 17

    In the dorsal SR, one-way ANOVA showed no significant effect of age on spine density [Figure 3e, F(4,90) 5 0.811, p 5 .52]. On the other
hand, one-way ANOVA revealed a significant effect of age on the proportion of spines in dorsal SR that were mature [Figure 3f, F(4,90) 5 5.959,
p 5 .0003]. Tukey’s HSD post hoc analysis showed that the proportion of spines that were mature was significantly higher at P44 (78.54 6 3.45%),
P50 (85.6 6 5.04%), and P55 (81.1 6 3.46%), compared with the P35 Group (61.547 6 2.34%; Figure 3f, all p < .05).
    This pattern for the dorsal SR was recapitulated in the dorsal SLM: one-way ANOVA showed no significant effect of age on the spine density [Fig-
ure 3g, F(4,90) 5 0.671, p 5 .61], while one-way ANOVA revealed a significant effect of age on the proportion of spines that were mature [Figure 3h, F
(4,90) 5 6.831, p < .0001]. Tukey’s HSD post hoc analysis showed that the proportion of mature spines was significantly higher at P50 (84.73 6 2.2%)
and P55 (87.41 6 2.62%), compared with P35 (63.841 6 4.82%; Figure 3h, p < .05) and P40 groups (69.47 6 3.55%; Figure 3h, p < .05).

                                                                    FIGURE 3.
8 of 17   |                                                                                                                              CHEN   ET AL.

3.5 | Dendritic spines of CA1 pyramidal cells in the ventral hippocampus mature with age and increase
in spine density
In ventral SR, the effect of age on dendritic spine density was different from that of dorsal CA1. Examples of dendritic segments from the SLM of
the ventral hippocampus at P35 and P55 groups are shown in Figure 3i and j, respectively. One-way ANOVA revealed a significant effect of age on
the dendritic spine density in the SR of ventral CA1 [Figure 3k, F(4,96) 5 7.313, p < .0001]. Tukey’s HSD post hoc analysis showed that the spine
density was significantly higher at P55 (1.79 6 0.1 spines per lm), compared with all other groups (P35: 1.34 6 0.09 spines per lm; P40:
1.34 6 0.08 spines per lm; P44: 1.07 6 0.08 spines per lm; P50: 1.22 6 0.07 spines per lm; Figure 3k, all p < .05).
    One-way ANOVA also showed a significant effect of age on the proportion of spines that were mature in the ventral SR [Figure 3l, F
(4,96) 5 3.489, p 5 .01]. Tukey’s HSD post hoc analysis showed that the proportion of spines that were mature was significantly higher at P50
(80.28 6 2.79%), compared with the P35 (67.81 6 3.01%; Figure 3l, p < .05) and P40 (64.05 6 3.11%; Figure 3l, p < .05) groups.
    In the ventral SLM, one-way ANOVA showed a significant effect of age on dendritic spine density [Figure 3m, F(4,100) 5 19.01,
p < .0001]. Tukey’s HSD post hoc analysis showed that the spine density was significantly higher at P50 (1.612 6 0.059 spines per lm)
and P55 (1.608 6 0.11 spines per lm), compared with P35 (1.289 6 0.066 spines per lm; Figure 3m, p < .05) and P40 (1.283 6 0.052
spines per lm; Figure 3m, p < .05). Surprisingly, Tukey’s HSD post hoc analysis also revealed that the spine density was significantly
lower at P44 (0.793 6 0.07 spines per lm), compared with all other groups (Figure 3m, all p < .05). The proportion of mature spines was
also significantly higher at P50 (76.97 6 2.9%), compared with the P35 (65.11 6 2.96%; Figure 3n, p < .05) and P44 (62.31 6 5.53%; Fig-
ure 3n, p < .05) groups.

3.6 | The effects of pair housing during adolescence are the opposite for dorsal versus ventral hippocampal CA1
Previously, we showed that environmental enrichment (wheel access) and stress (food restriction) elicit pathway-specific changes upon CA1 pyrami-
dal neurons in the dorsal and ventral hippocampus of adolescent female rats (Chowdhury et al., 2014a). In that study, all animals had been singly-
housed. Thus, it remained to be determined whether pair housing during adolescence also causes dendritic remodeling in the hippocampal CA1. Tis-
sue from another cohort of 16 adolescent rats was used to address this question.
    Dorsal CA1 pyramidal neurons’ apical dendrites in pair-housed animals had overall marginally significantly greater number of dendritic intersec-
tions with Sholl spheres, relative to those from singly-housed controls [Figure 4a, F(1,60) 5 1.724, p 5 .19]. The average number of intersections in
the SR was significantly greater in the pair-housed group (4.673 6 0.359) than in the singly-housed group [3.725 6 0.281; Figure 4b; t(60) 5 2.095,
p 5 .04]. There was a trend for increased average dendritic lengths in the SR of pair-housed animals (615.753 6 51.74 lm), compared with singly-
housed controls [516.45 6 40.93 lm; Figure 4c; t(60) 5 1.515, p 5 .13].
    In contrast to dorsal SR, dorsal SLM showed no difference in the number of intersections or dendritic lengths between singly-housed (intersec-
tions: 2.056 6 0.25; length: 441.2 6 46.37 lm) and pair-housed animals [intersections: 1.913 6 0.27; Figure 4d; t(60) 5 0.389, p 5 0.7; length:
444.8 6 46.36 lm; Figure 4e; t(32) 5 0.056, p 5 .96].

FIGURE 3       Dendritic spine analysis reveals that the proportion of spines that are mature increases with age in the dorsal hippocampus,
while spine density increases with age in the ventral hippocampus. Left panels (a) and (b) show the higher magnification of the
representative projection of Golgi-stained apical dendritic segments in the SR of the dorsal hippocampus from the P55 group [panel (d)] and
apical dendritic segments in the SLM of the ventral hippocampus from P55 group [panel (j)], created by joining portions in different planes
of focus. White triangles indicate the specific spine showed in panel (a) and (d) (immature) and in panels (b) and (j) (mature). In the right pan-
els (a) and (b), Neurolucida tracings of the same dendritic fragments are showed. Filled circles represent spines categorized as mature (mush-
room or stubby spines). Empty circles represent spines categorized as immature (thin or filopodia spines). Calibration bar represents 2 mm.
Representative projection of Golgi-stained apical dendritic segments in SR of the dorsal hippocampus from the P35 group [(c), top panel]
and from the P55 group [(d), top panel], created by joining portions in different planes of focus. In the bottom of panels (c) and (d), Neurolu-
cida tracings of the same dendritic fragments are shown. The calibration bar represents 10 mm. Density and proportion of mature spines in
the SR and the SLM were calculated for each dendritic segment within each age group. In the dorsal hippocampal CA1, total spine density
measured in the SR (e) and in the SLM (g) showed no significant difference across the ages. The proportion of mature spines in the SR (f)
and the SLM (h) increased with age. Representative projection of Golgi-stained apical dendritic segments in SLM of the ventral hippocampus
from P35 group [(i), top panel], and from P55 group [(j), top panel], created by joining portions in different planes of focus. In bottom panels
(i) and (j), Neurolucida tracings of the same dendritic segments are showed. Calibration bar represents 10 mm. In the ventral hippocampal
CA1, total spine density measured in the SR (k) and in the SLM (m) increased with age. The proportion of mature spines in the SR (l) and
the SLM (n) increase transiently, with the peak at P50. Data from each group were compared across ages from P35 to P55. Bar graphs rep-
resent means 1 SEM (N 5 15–31 fragments per group). Same letters indicate significant differences between age groups as assessed by
one-way ANOVA, followed by Tukey’s HSD test. The asterisk indicates significant difference across combined age groups, as assessed by
the same statistical tests
CHEN   ET AL.                                                                                                                          |   9 of 17

FIGURE 4       Pair housing during early to mid-adolescence has opposite effects on apical dendritic branching in dorsal versus ventral hippo-
campal CA1. The number of intersections with Sholl spheres centered at the soma of pair-housed animals’ and singly-housed controls’ dorsal
(a) and ventral (f) CA1 at P44 are plotted against radii of the sphere as the means 1 SEM (dorsal: N 5 30 cells for pair-housed, 32 cells for
singly-housed group; ventral: N 5 37 cells for pair-housed, 35 cells for singly-housed group). In the dorsal hippocampal CA1, pair housing
during adolescence increases the average number of dendritic intersections in the SR (b), but not in the SLM (d). There is also a trend
toward an increase in total dendritic lengths in the SR (c), but not in the SLM (e). In ventral hippocampal CA1, pair housing during adoles-
cence decreases the average number of dendritic intersections in the SR (g), but not in the SLM (i). No significant difference between
groups was found for total dendritic lengths in the SR (h) or the SLM (j). S, singly-housed controls. P, pair-housed animals. Asterisks indicate
statistical significance with p < .05 by Student’s t tests
10 of 17   |                                                                                                                             CHEN   ET AL.

    Sholl analysis of neurons from the ventral hippocampus showed effects of pair housing that were the opposite of those observed in the
dorsal hippocampus: pyramidal neurons of the pair-housed group exhibited marginally significantly fewer number of dendritic intersections
with Sholl spheres, relative to those of singly-housed controls [Figure 4f; F(1,70) 5 3.082, p 5 .084]. The average number of intersections in
the SR was significantly smaller in the pair-housed group (3.0 6 0.26) than in singly-housed controls [3.806 6 0.29; Figure 4g; t(70) 5 2.088,
p 5 .04]. There was also a trend for decreased average dendritic lengths in the ventral SR of pair-housed animals (431.87 6 43.77 lm), com-
pared with singly-housed controls [520.13 6 37.14 lm; Figure 4h; t(70) 5 1.529, p 5 .13]. No difference was found in the number of inter-
sections or dendritic lengths in the SLM of the ventral CA1 between the pair-housed group (intersections: 1.724 6 0.17; length:
298.9 6 21.16 lm) and singly-housed controls [intersections: 1.434 6 0.18; Figure 4i, t(70) 5 1.167, p 5 .25; length: 281.1 6 32.69 lm; Fig-
ure 4j, t(46) 5 0.475, p 5 .64].

3.7 | Lamina-specific effects of pair housing during adolescence on dendritic spine density of the dorsal
and ventral hippocampal CA1 neurons
Spine density analysis was conducted to study the effects of pair housing during adolescence on overall spine density and proportion of mature
spines. Examples of dendritic segments from the SR of the dorsal hippocampus of singly-housed and pair-housed animals are shown in Figure 5a
and b, respectively. SR of the dorsal and ventral hippocampal CA1 showed no difference between pair-housed and singly-house groups in terms of
spine density (Figure 5c and g) or the proportion of spines that were mature (Figure 5d and h). On the other hand, in SLM of both dorsal and ventral
hippocampal CA1, spine density was significantly increased among neurons in the pair-housed group, compared with neurons from the singly-
housed group [dorsal: Figure 5e, t(34) 5 3.28, p 5 .002; ventral: Figure 5i, t(29) 5 2.22, p 5 .03]. No difference between groups was found regard-
ing the proportion of spines that were mature in SLM of dorsal or ventral hippocampal CA1 (Figure 5f and j).

4 | DISCUSSION

To the best of our knowledge, this is the first systematic study of the development of pyramidal neurons in the hippocampal CA1 of female rats dur-
ing adolescence. By sampling neurons at multiple time points, we were able to capture global retractions and growth spurts during adolescence that
would have been missed with fewer time points. We demonstrate that during adolescence, the dendritic arbor and spines of pyramidal cells in the
dorsal and ventral hippocampal CA1 are remodeled differently. Figure 6 summarizes the structural changes of dendritic lengths, branching complex-
ity, and spine density, ultimately providing an estimate of spine number (spine density 3 dendritic length) in the dorsal and ventral CA1 neurons
from early-, mid-, and late-adolescence.
    The dorsal hippocampi of singly-housed female rats exhibit transient global retractions during early- to mid-adolescence, followed by protracted
maturation, evident as a 55% growth spurt of dendritic branching and 49% growth spurt in total dendritic lengths between P50 and P55. The tran-
sient decrease of dendritic branching from P35 to P50 is consistent with a previous finding that the total number of dendritic spines decreases dur-
ing adolescence from P35 to P49 in female rats (Yildirim et al., 2008).
    CA1 pyramidal cells of the ventral hippocampi of singly-housed female rats exhibit earlier, transient lamina-specific growth during adolescence,
since dendrites of increase in branching in the SR between the adolescent ages of P44 and P50, then retract toward late adolescence/early adult-
hood (P55), with no concomitant change in the SLM. These findings of growth, followed by retraction of dendritic complexity in SR of the ventral
hippocampus from mid- to late-adolescence, are consistent with our previous finding (Chowdhury et al., 2014b).

4.1 | Dendrites spines of hippocampal CA1 neurons exhibit developmental changes that are different
for the dorsal versus ventral sectors
Dendritic spine density and maturity in the hippocampal CA1 also vary with age. Dendritic spines in the dorsal hippocampus follow a predictable
pattern. They mature with age but do not change in density. This suggests transitioning of existing spines from newly formed immature types to rel-
atively mature stable types from P35 to P44 toward P55. However, because of the transient decrease in dendritic branching and lengths at mid-
adolescence (P50), total spine number in SR is estimated to also undergo a transient decrease at this time. What pivotal event may be associated
with mid-adolescence is an interesting and important question for future studies.
    The observed constant spine density during adolescence differs from the pruning in spine density reported previously for adolescent female
mice (Afroz et al., 2016). We focused on the apical dendritic tree of the pyramidal neurons, while Afroz and colleagues pooled the segments from
both basal and apical dendrites and did not differentiate measurements in the dorsal versus ventral hippocampus. The difference between the two
studies suggests that the relative influence of different afferent inputs on basal and apical dendritic trees may change during adolescent develop-
ment, with pruning occurring within the basal dendrites (part of data of Afroz et al.,) and no pruning of apical dendrites (our data).
    In the ventral hippocampus, dendritic spines change during adolescence in a less predictable way. Dendritic density increases during late adoles-
cence, but there is also a transient enrichment of mature spines during mid-adolescence. These findings suggest the addition of immature spines
CHEN   ET AL.                                                                                                                              |   11 of 17

FIGURE 5      Pair housing during adolescence increases dendritic spine density in the SLM of dorsal and ventral hippocampal CA1.
Representative projection of Golgi-stained apical dendritic fragments in the SLM of the dorsal hippocampus from the singly-housed control
[(a) left panel] and from the pair-housed group [(b) left panel], created by joining portions in different planes of focus. In right panels (a) and
(b), Neurolucida tracings of the same dendritic fragments are shown. Filled circles represent spines categorized as mature (mushroom or
stubby spines). Empty circles represent spines categorized as immature (thin spines or filopodia). Calibration bar represents 10 mm. Density
and proportion of mature spines in the SR and the SLM were calculated for each dendritic fragment within each group. In both dorsal and
ventral hippocampal CA1, total spine density (mature 1 immature) was not different between the groups in the SR [dorsal: panel (c); ventral:
panel (g)], but was greater in the SLM of pair-housed animals [dorsal: panel (e); ventral: panel (i)]. The proportion of spines that were mature
was not significantly different across the groups in the SR [dorsal: panel (d); ventral: panel (h)] or the SLM [dorsal: panel (f); ventral: panel
(j)]. Bar graphs represent means 1 SEM (N 5 16–33 fragments per group). S, singly-housed controls. P, pair-housed animals. Asterisks indi-
cate statistically significant difference with p < .05 by Student’s t tests

during late adolescence. Since late adolescence is when dendrites are also lengthening, there is a significant net increase of spines, including imma-
ture ones, during late adolescence. What environmental, endocrine or afferent activity factors may contribute to the late adolescent phase of spine
addition in SR of the ventral hippocampus is another interesting question for future studies.
    One potential factor driving spine changes in the dorsal hippocampus at mid-adolescence and in the ventral hippocampus at late adolescence
could be the estrogen receptor alpha (ERa) and G-protein-coupled estrogen receptor type 1 (GPER1): both of these have been demonstrated to
occur in dendritic spines of the hippocampus, including those in SR of dorsal CA1, and to participate in synaptic plasticity (McEwen et al. 2001;
Waters et al., 2015). GPER1 shows a dorsal-ventral distinction in its effects on cell proliferation in the hippocampus of adult female rats (Duarte-
Guterman, Lieblich, Chow, & Galea, 2015). The hippocampus expresses ERa and b from prenatal stages (Gerlach, McEwen, Toran-Allerand, & Fried-
man, 1983; Gonzalez et al., 2007) through adulthood (Gonzalez et al., 2007; Hart, Patton, & Woolley, 2001; McEwen et al., 2001; Scudiero & Ver-
derame, 2017) and with contrasting patterns across pyramidal versus GABAergic neurons in the dorsal versus ventral sectors (Hart et al., 2001). It
remains unexplored whether the expression of any of these receptors exhibit transient rises and dips across cell types or across dorsal-to-ventral
sectors of the hippocampus during adolescence.
12 of 17   |                                                                                                                               CHEN   ET AL.

FIGURE 6     Summary of lamina-specific changes in dendrites of the dorsal and ventral hippocampal CA1 pyramidal neurons during adoles-
cent development (right-ward black arrows) and dendritic remodeling evoked by pair housing during adolescence (up- and downward
arrows). (1) Pyramidal neurons of the dorsal hippocampus transiently decrease in dendritic complexity, while pyramidal neurons of the ven-
tral hippocampus transiently increase in dendritic complexity during adolescence. (2) Dendritic spines in the dorsal hippocampus mature
with age but do not change in density. In the ventral hippocampus, dendritic spine density increases during adolescence and exhibits tran-
sient enrichment of mature spines during mid-adolescence. (3) Pair housing during early to mid-adolescence (gray arrows) increase apical
dendritic branching in the dorsal hippocampus, but decrease dendritic branching in the ventral hippocampus. (4) Pair housing during adoles-
cence increases dendritic spine density specifically in the SLM of both dorsal and ventral hippocampal CA1

4.2 | Dendrites of the dorsal and ventral hippocampal CA1 neurons exhibit transient lamina-specific growths
and retractions during adolescence
The hippocampus is a component of the limbic system that plays a role in spatial learning, memory and the regulation of stress (Bannerman et al.,
2003; Henke, 1990; McCormick et al., 2010; Moser, Moser, Forrest, Andersen, & Morris, 1995). The hippocampus undergoes robust development
during the onset of puberty and with the introduction of the gonadal hormones (Andersen & Teicher, 2004; Zitman and Richter-Levin, 2013).
    An early introduction to stress can cause structural changes in the hippocampus (Buwalda, Geerdink, Vidal, & Koolhaas, 2011; Salas-Ramirez,
Frankfurt, Alexander, Luine, & Friedman, 2010) such as the size of the hippocampus (Leussis & Andersen, 2008) and dendritic remodeling of the api-
cal CA1 pyramidal cells (Chowdhury et al., 2014a,b; Isgor, Kabbaj, Akil, & Watson, 2004). Changes in the levels of glucocorticoid receptors in the hip-
pocampus are also associated with changes in the morphology, either increasing or decreasing the length and branching in the dendrites of the
CHEN   ET AL.                                                                                                                                    |   13 of 17

pyramidal neurons (Andersen & Teicher, 2004; Loi, Koricka, Lucassen, & Joels, 2014; Toledo-Rodriguez & Sandi, 2011). In fact, our observations of
transient growth and retraction in the dendritic branching during adolescence, with peak and dip around P44 to P50, echo the results from previous
findings, namely that the stress-induced corticosterone response is lowest at P50 during adolescence (Foilb, Lui, & Romeo, 2011; Romeo, 2013).
    The dorsal hippocampus forms a cortical network with the retrosplenial and anterior cingulate cortical areas that connect to other sensorimotor
cortical areas and mediate cognitive processes, such as learning, memory, navigation, and exploration (Fanselow & Dong, 2010). The ventral hippo-
campus is connected with the amygdalar nuclei, bed nucleus of the stria terminalis, olfactory bulb, lateral septum, nucleus accumbens, and infralimbic
and prelimbic areas of the medial prefrontal cortex (Adhikari, Topiwala, & Gordon, 2010; Adhikari, Topiwala, & Gordon, 2011; Fanselow & Dong,
2010). Ultimately, all of these structures innervate the hypothalamus, thus mediating motivated behaviors with strong emotional components. The
ventral hippocampus receives connections from the medial division of the entorhinal cortex, which receives olfactory, gustatory, and visceral inputs
(Fanselow & Dong, 2010). Thus, the transient growth and retraction of dendritic branching in the hippocampus during adolescence may contribute
to the instability of emotional regulation and control of cognitive behaviors during adolescence (Geier, 2013). The U-shaped developmental trajec-
tory of dorsal hippocampal CA1 dendrites may contribute to the reduced ability of cognitive control and increased impulsivity during mid-
adolescence. The inverted U-shaped developmental curve in the ventral hippocampal CA1 dendrites may contribute to the particularly strong reac-
tivity in the form of anxiety that adolescents exhibit in response to stress (Dahl, 2004; Romeo, 2010).

4.3 | Pair housing causes morphological changes during puberty across hippocampal regions
Our data show that pair housing during early to mid-adolescence increases apical dendritic branching of pyramidal neurons in the dorsal hippocam-
pus, but decreases apical dendritic complexity of pyramidal neurons in the ventral hippocampus. These effects could be due to greater sensory stim-
ulation and/or additional factors associated with pair housing. Previous studies also reported enhanced dendritic growth and branching in layer III of
parietal cortex of puberty rats reared in enriched environments consisting of group housing, shelter, plastic toys, and a running wheel (Leggio et al.,
2005). On the other hand, previous studies of neuronal development have concentrated on sensory cortices, and a large body of evidence exists
describing dendritic remodeling following sensory deprivation during the juvenile critical period that is well before puberty (Antonini, Fagiolini, &
Stryker, 1999; Fagiolini, Pizzorusso, Berardi, Domenici, & Maffei, 1994; Hensch & Stryker, 2004; Kirkwood, Lee, & Bear, 1995; Valverde, 1967). Not
only has increased spine density been found in adolescent and adult rodent brains following rearing in an enriched environment (Globus, Rose-
nzweig, Bennett, & Diamond, 1973; Jung & Herms, 2014), but there are also reports of transient increases in spine number and in the turnover rate
of spines in adulthood (Jung & Herms, 2014). Furthermore, previous studies also report that post-weaning social isolation upon males induces reduc-
tions in dendritic spine density on hippocampal pyramidal neurons (Silva-Gomez, Rojas, Juarez, & Flores, 2003), enhancements in hippocampal
potassium ion channel currents (Quan, Tian, Xu, Zhang, & Yang, 2010), and alterations in hippocampal cell proliferation (McCormick et al., 2010).
Our study adds to this large body of literature by showing that enrichment of adolescent females in the form of pair housing has influences on den-
dritic remodeling.
    We have previously shown that food restriction, combined with social isolation, contributed to a dorsal–ventral distinction of changes in dendritic
branching of CA1 pyramidal neurons (Chowdhury et al., 2014a). Our current results further indicate that apical dendritic structure of CA1 pyramidal neu-
rons is modifiable during adolescence with less environmental perturbation (just isolation, no food restriction). Moreover, the direction of changes in api-
cal dendritic structure under pair housing is the opposite between dorsal and ventral hippocampus, which may be related to the distinct pattern of
connectivity of the hippocampal subregions with other brain regions, as was discussed in the previous section. Thus, the increase in dendritic branching
in dorsal hippocampal CA1 may be due to increased sensory stimulation by pair housing. On the other hand, in ventral hippocampal CA1, the decrease
in dendritic branching may be attributable to the stress-reducing effect of pair housing that leads to less synaptic inputs from the amygdala.
    Within a single CA1 pyramidal neuron, the apical dendrites integrate information arriving from several different sources. Therefore, remodeling
of dendrites in one particular layer may imply the importance of inputs arriving by that particular pathway, within specific laminae. The increased
dendritic branching in the SR of dorsal CA1 of pair-housing animals may be caused by increased probability of these pyramidal neurons to receive
inputs from neighboring CA1 pyramidal neurons and CA3 pyramidal cells. In the ventral CA1, our data indicate a decrease in dendritic branching in
the SR with pair housing, which may be caused by decreased afferent activity in the SR, arising from the entorhinal cortex or basal nucleus of the
amygdala (Pikkarainen et al., 1999), in addition to the neighboring CA1 and CA3 pyramidal neurons. While in the SR of the dorsal and ventral CA1
showed changes in response to pair housing, there was no difference in dendritic branching of the SLM across the rearing groups. This suggests that
the pair housing-evoked plasticity was more strongly influenced by afferents arising from the CA3, neighboring CA1 cells, or local inhibition to the
SR dendrites than by the entorhinal cortical inputs to the SLM. However, pair housing during adolescence increases dendritic spine density specifi-
cally in the SLM of both dorsal and ventral hippocampal CA1, which might suggest an increase in the ratio between excitation and inhibition in the
SLM of both dorsal and ventral hippocampal CA1 regions that receive inputs from the entorhinal cortical areas. SLM is the recipient zone of excita-
tory inputs to the thalamic reuniens nucleus (Herkenham, 1978). This pathway is important for mediating the communication in the prefrontal
cortex-to-hippocampus direction (Roy, Svensson, Mazeh, & Kocsis, 2017) as a return pathway to the more well-studied connection from the ventral
hippocampus to the prefrontal cortex that mediate regulation of anxiety (Adhikari et al., 2010, 2011).
14 of 17   |                                                                                                                                     CHEN   ET AL.

4.4 | Closing remarks
In this study, we showed that the dorsal and ventral hippocampi of singly-housed female animals exhibit layer-specific changes through develop-
ment, evident as substantial and dynamic remodeling within the 20 days spanning early-, mid-, and late- adolescence (P35–P55). These findings indi-
cate that pyramidal cells in the dorsal and the ventral hippocampal CA1 are remodeled profoundly and differently in ways that might contribute to
variability in emotional regulations and cognitive controls during adolescence. We also show that pair housing with another female from P36 to 44
significantly increases branching in the SR of dorsal hippocampus and reduces the branching in the SR of ventral hippocampus in female rats, which
suggests that pair housing during adolescence elicits pathway-specific changes in the hippocampus that may underlie the behavioral benefits
observed in social housing. Taken together with data on spine density, these results strengthen the hypothesis that the hippocampus is particularly
sensitive to experience-dependent environmental changes during adolescence. Previous study indicated that the adolescent trait of impulsivity is
normalized by enriched rearing during adolescence (Adriani et al., 2006). Our previous data also suggest a dorsal–ventral distinction of changes in
dendritic branching elicited by food restriction stress after puberty (Chowdhury et al., 2014a), and food restriction experience during adolescence is
beneficial for cognitive performances in early adulthood (Chowdhury, Fenton, & Aoki, 2017, Under review). To understand how pair housing and
the anatomical changes in the SR and SLM of the dorsal and ventral hippocampal CA1 contribute to behaviors along the developmental trajectory
of adolescence, it is necessary to examine how neuroanatomical changes in afferent and efferent pathways of the hippocampus CA1 impact func-
tion. Our future studies will incorporate electrophysiology with behavioral testing at early-, mid-, and late-adolescence to explore how functional
changes in circuits are involved in the hippocampus-dependent behaviors.

AC KNOW LE DGME NT S
This study was supported by The Klarman Family Foundation Grant Program in Eating Disorders Research, National Institutes of Health, Grant
Number: R21MH091445-01; Grant Number: R21 MH105846; Grant Number: R01NS066019-01A1; Grant Number: R01NS047557-07A1;
Grant Number: NEI Core grant EY13079; NYUs Research Challenge Fund, Grant Number: NSF-REU 1460880 to CA, YWC, and the Fulbright
Scholarship to YWC. National Center for the Advancing of Translational Science (NCATS), Grant Number: UL1 TR000038 to TGC. National
Institutes of Health, Grant Number: R25GM097634-01 to RA. We thank Walter Piper for his assistance with the initial phase of the experi-
ment. We thank Wei Huang, Lauren Klingensmith, Anuj Rao, Aja Evans, Faith Shaeffer, and Hope Balsmeyer for their technical assistance of
the Golgi staining steps. We thank Dr. Keith Purpura for the helpful discussion revolving statistical analysis.

C ONFLICT OF INT E RE ST

We declare no conflict of interest in relation with the work described.

OR CID
Chiye Aoki     http://orcid.org/0000-0003-4010-9425

R E FER E NCE S
Adhikari, A., Topiwala, M. A., & Gordon, J. A. (2010). Synchronized activity between the ventral hippocampus and the medial prefrontal cortex during
   anxiety. Neuron, 65(2), 257–269.
Adhikari, A., Topiwala, M. A., & Gordon, J. A. (2011). Single units in the medial prefrontal cortex with anxiety-related firing patterns are preferentially
   influenced by ventral hippocampal activity. Neuron, 71(5), 898–910.
Adriani, W., Giannakopoulou, D., Bokulic, Z., Jernej, B., Alleva, E., & Laviola, G. (2006). Response to novelty, social and self-control behaviors, in rats
   exposed to neonatal anoxia: Modulatory effects of an enriched environment. Psychopharmacology (Berlin), 184(2), 155–165.
Afroz, S., Parato, J., Shen, H., & Smith, S. S. (2016). Synaptic pruning in the female hippocampus is triggered at puberty by extrasynaptic GABAA recep-
   tors on dendritic spines. eLife, 5, e15106. https://doi.org/10.7554/eLife.15106
Amaral, D., L. P. (2007). Hippocampal neuroanatomy. In P. Andersen, R. Morris, D. Amaral, T. Bliss, & J. O’Keefe (Eds.), The hippocampus book (pp. 37–
  114). New York, Oxford University Press.
Andersen, S. L., & Teicher, M. H. (2004). Delayed effects of early stress on hippocampal development. Neuropsychopharmacology, 29(11), 1988–1993.
Andersen, S. L., Thompson, A. T., Rutstein, M., Hostetter, J. C., & Teicher, M. H. (2000). Dopamine receptor pruning in prefrontal cortex during the peri-
   adolescent period in rats. Synapse, 37(2), 167–169.
Antonini, A., Fagiolini, M., & Stryker, M. P. (1999). Anatomical correlates of functional plasticity in mouse visual cortex. Journal of Neuroscience, 19(11),
   4388–4406.
Aoki, C., Chen, Y. W., Chowdhury, T. G., & Piper, W. (2017). a4bd-GABAA receptors in dorsal hippocampal CA1 of adolescent female rats traffic to the
   plasma membrane of dendritic spines following voluntary exercise and contribute to protection of animals from activity-based anorexia through
   localization at excitatory synapses. Journal of Neuroscience Research. https://doi.org/10.1002/jnr.24035
CHEN   ET AL.                                                                                                                                        |   15 of 17

Bannerman, D. M., Deacon, R. M., Offen, S., Friswell, J., Grubb, M., & Rawlins, J. N. (2002). Double dissociation of function within the hippocampus:
   Spatial memory and hyponeophagia. Behavioural Neuroscience, 116(5), 884–901.
Bannerman, D. M., Grubb, M., Deacon, R. M., Yee, B. K., Feldon, J., & Rawlins, J. N. (2003). Ventral hippocampal lesions affect anxiety but not spatial
   learning. Behavioural Brain Research, 139(1–2), 197–213.
Bayer, S. A., & Altman, J. (1974). Hippocampal development in the rat: Cytogenesis and morphogenesis examined with autoradiography and low-level
   X-irradiation. Journal of Comparative Neurology, 158(1), 55–79.
Buwalda, B., Geerdink, M., Vidal, J., & Koolhaas, J. M. (2011). Social behavior and social stress in adolescence: A focus on animal models. Neuroscience
   Biobehavior & Review, 35(8), 1713–1721.
Casey, B. J., Jones, R. M., & Hare, T. A. (2008). The adolescent brain. Annals of the New York Academy of Sciences, 1124, 111–126.
Casey, B. J., Tottenham, N., Liston, C., & Durston, S. (2005). Imaging the developing brain: What have we learned about cognitive development? Trends
   in Cognitive Science, 9(3), 104–110.
Cavdar, S., Onat, F. Y., Cakmak, Y. O., Yananli, H. R., Gulcebi, M., & Aker, R. (2008). The pathways connecting the hippocampal formation, the thalamic
   reunions nucleus and the thalamic reticular nucleus in the rat. Journal of Anatomy, 212(3), 249–256.
Chowdhury, T. G., Barbarich-Marsteller, N. C., Chan, T. E., & Aoki, C. (2014). Activity-based anorexia has differential effects on apical dendritic branch-
   ing in dorsal and ventral hippocampal CA1. Brain Structure & Function, 219(6), 1935–1945.
Chowdhury, T. G., Fenton, A. A., & Aoki, C. (2017, Under review). The effects of adolescent experience of food restriction and exercise on spatial learn-
   ing and open field exploration.
Chowdhury, T. G., Rios, M. B., Chan, T. E., Cassataro, D. S., Barbarich-Marsteller, N. C., & Aoki, C. (2014). Activity-based anorexia during adolescence
   disrupts normal development of the CA1 pyramidal cells in the ventral hippocampus of female rats. Hippocampus, 24(12), 1421–1429.
Clark, I., & Dumas, G. (2015). Toward a neural basis for peer-interaction: What makes peer-learning tick? Frontiers in Psychology, 6, 28.
Conrad, C. D., Ortiz, J. B., & Judd, J. M. (2017). Chronic stress and hippocampal dendritic complexity: Methodological and functional considerations.
   Physiological Behaviour, 178, 66–81.
Cooke, B. M., & Woolley, C. S. (2005). Gonadal hormone modulation of dendrites in the mammalian CNS. Journal of Neurobiology, 64(1), 34–46.
Dahl, R. E. (2004). Adolescent brain development: A period of vulnerabilities and opportunities. Keynote address. Annals of the New York Academy of
   Sciences, 1021(1), 1–22.
Drzewiecki, C. M., Willing, J., & Juraska, J. M. (2016). Synaptic number changes in the medial prefrontal cortex across adolescence in male and female
   rats: A role for pubertal onset. Synapse, 70(9), 361–368.
Duarte-Guterman, P., Lieblich, S. E., Chow, C., & Galea, L. A. (2015). Estradiol and GPER activation differentially affect cell proliferation but not GPER
   expression in the hippocampus of adult female rats. PLoS One, 10(6), e0129880.
Fagiolini, M., Pizzorusso, T., Berardi, N., Domenici, L., & Maffei, L. (1994). Functional postnatal development of the rat primary visual cortex and the
   role of visual experience: Dark rearing and monocular deprivation. Vision Research, 34(6), 709–720.
Faherty, C. J., Kerley, D., & Smeyne, R. J. (2003). A Golgi–Cox morphological analysis of neuronal changes induced by environmental enrichment. Brain
   Research & Development, 141(1–2), 55–61.
Fanselow, M. S., & Dong, H. W. (2010). Are the dorsal and ventral hippocampus functionally distinct structures? Neuron, 65(1), 7–19.
Foilb, A. R., Lui, P., & Romeo, R. D. (2011). The transformation of hormonal stress responses throughout puberty and adolescence. Journal of Endocrinol-
    ogy, 210(3), 391–398.
Galef, B. G., & Sorge, R. E. (2000). Use of PVC conduits by rats of various strains and ages housed singly and in pairs. Journal of Applied Animal Welfare
   Science, 3(4), 279–292.
Geier, C. F. (2013). Adolescent cognitive control and reward processing: Implications for risk taking and substance use. Hormonal Behaviour, 64(2), 333–342.
Gerlach, J. L., McEwen, B. S., Toran-Allerand, C. D., & Friedman, W. J. (1983). Perinatal development of estrogen receptors in mouse brain assessed by
   radioautography, nuclear isolation and receptor assay. Brain Research, 11(1), 7–18.
Giedd, J. N. (2004). Structural magnetic resonance imaging of the adolescent brain. Annals of the New York Academy of Sciences, 1021, 77–85.
Giedd, J. N., Clasen, L. S., Lenroot, R., Greenstein, D., Wallace, G. L., Ordaz, S., . . . Chrousos, G. P. (2006). Puberty-related influences on brain develop-
   ment. Molecular Cell & Endocrinology, 254–255, 154–162.
Globus, A., Rosenzweig, M. R., Bennett, E. L., & Diamond, M. C. (1973). Effects of differential experience on dendritic spine counts in rat cerebral cor-
   tex. Journal of Comprehensive Physiology & Psychology, 82(2), 175–181.
Gogtay, N., Giedd, J. N., Lusk, L., Hayashi, K. M., Greenstein, D., Vaituzis, A. C., . . . Thompson, P. M. (2004). Dynamic mapping of human cortical devel-
   opment during childhood through early adulthood. Proceedings of National Academic Sciences USA, 101(21), 8174–8179.
Gonzalez, M., Cabrera-Socorro, A., Pe
                                     rez-García, C. G., Fraser, J. D., Lo
                                                                          pez, F. J., Alonso, R., & Meyer, G. (2007). Distribution patterns of estrogen recep-
   tor alpha and beta in the human cortex and hippocampus during development and adulthood. Journal of Comparative Neurology, 503(6), 790–802.
Harris, K. M., & Kater, S. B. (1994). Dendritic spines: Cellular specializations imparting both stability and flexibility to synaptic function. Annual Review of
   Neuroscience, 17(1), 341–371.
Hart, S. A., Patton, J. D., & Woolley, C. S. (2001). Quantitative analysis of ER alpha and GAD colocalization in the hippocampus of the adult female rat.
   Journal of Comparative Neurology, 440(2), 144–155.
Hazen, E., Schlozman, S., & Beresin, E. (2008). Adolescent psychological development: A review. Pediatric Review, 29(5), 161–167; quiz 168.
You can also read